thyroid hormone action in the adult brain: gene …digital.csic.es/bitstream/10261/22145/1/thyroid...

Thyroid Hormone Action in the Adult Brain: GeneExpression Profiling of the Effects of Single and MultipleDoses of Triiodo-L-Thyronine in the Rat Striatum

Diego Diez, Carmen Grijota-Martinez, Patrizia Agretti, Giuseppina De Marco, Massimo Tonacchera,Aldo Pinchera, Gabriella Morreale de Escobar, Juan Bernal, and Beatriz Morte

Instituto de Investigaciones Biomedicas (D.D., C.G.-M., G.M.d.E., J.B., B.M.), Consejo Superior de InvestigacionesCientıficas and Universidad Autonoma de Madrid (CSIC-UAM), and Centro de Investigacion Biomedica en Red deEnfermedades Raras (CIBERER) (C.G.-M., G.M.d.E., J.B., B.M.), Instituto de Salud Carlos III, 28029 Madrid, Spain; andDepartment of Endocrinology and Metabolism (P.A., G.D.M., M.T., A.P.), Centro Eccellenza AmbiSEN, University of Pisa,56124 Pisa, Italy

Thyroid hormones have profound effects on mood and behav-ior, but the molecular basis of thyroid hormone action in theadult brain is relatively unknown. In particular, few thyroidhormone-dependent genes have been identified in the adultbrain despite extensive work carried out on the developingbrain. In this work we performed global analysis of gene ex-pression in the adult rat striatum in search for genomicchanges taking place after administration of T3 to hypothy-roid rats. The hormone was administered in two differentschedules: 1) a single, large dose of 25 �g per 100 g body weight(SD) or 2) 1.5 �g per 100 g body weight once daily for 5 d (RD).Twenty-four hours after the single or last of multiple doses,

gene expression in the striatum was analyzed using Codelinkmicroarrays. SD caused up-regulation of 149 genes and down-regulation of 88 genes. RD caused up-regulation of 18 genesand down-regulation of one gene. The results were confirmedby hybridization to Affymetrix microarrays and by TaqManPCR. Among the genes identified are genes involved in cir-cadian regulation and the regulation of signaling pathwaysin the striatum. These results suggest that thyroid hormoneis involved in regulation of striatal physiology at multiplecontrol points. In addition, they may explain the beneficialeffects of large doses of thyroid hormone in bipolar disorders.(Endocrinology 149: 3989–4000, 2008)

THE EFFECTS AND mechanisms of action of thyroidhormone on the developing brain have been exten-

sively studied (1, 2). Thyroid hormone has multiple actionson brain maturation, resulting from regulation of neuralcell migration and differentiation, synaptogenesis, andmyelination. Hypothyroidism alters neuronal migrationin the cerebral cortex and the cerebellum and impairs thedifferentiation of pyramidal cells, Purkinje cells, �-ami-nobutyric acid (GABA)ergic cell precursors, oligodendro-cytes, astrocytes, and microglia. Most if not all of theseactions of thyroid hormone are due to regulation of geneexpression mediated by interaction with transcriptionallyactive nuclear receptors (3).

The molecular basis of thyroid hormone action in the ma-ture brain is less well known, despite its importance in brainfunction. In adults thyroid hormone influences mood andbehavior, and thyroid dysfunction very often leads to psy-chiatric disorders (4). High doses of T4 are effective in bipolardepression (5, 6). Despite these observations, very little isknown on the mechanisms of action of thyroid hormone inthe adult brain. Neurotransmitter systems are affected by

deficiency or excess of thyroid hormones (7), and thyroidhormone influences neurogenesis in the subventricular andsubgranular zones in the adult rat brain (8, 9). Thyroid hor-mone receptors are widely expressed in the adult brain, andparticularly the TR�1 isoform has been implicated in thecontrol of pathways regulating behavior (10). As in othertissues, it is most likely that the action of thyroid hormone inthe adult brain is exerted through the control of gene ex-pression. However, an important feature of thyroid hormoneaction is that the genes regulated by thyroid hormones in thedeveloping brain are insensitive to thyroid hormone in theadult brain, with some exceptions (2).

The purpose of the present work was to fill an importantgap in our knowledge of thyroid hormone action in the adultbrain by the identification of thyroid hormone-responsivegenes. For this study we selected the striatum for two mainreasons. One reason was because the thyroid hormone reg-ulated gene Nrgn (also known as RC3), is responsive to T3 inthe striatum but not other regions of the adult brain (11).Another reason was the problems of sensitivity thresholddue to the high complexity of the whole brain: the cellularcomplexity of the striatum is not as high as other regions ofpotential interest such as the cerebral cortex because a singlecell type, the GABAergic medium-spiny projection neuron,represents more than 90% of the total cell population. It isworth mentioning in this context that the striatum was oneof the regions showing altered metabolic activity after ad-ministration of T4 to bipolar patients (6).

We analyzed the effects of single and multiple doses of T3

First Published Online May 8, 2008Abbreviations: DARPP, Dopamine and cAMP-regulated phospho-

protein; D1, type 1 deiodinase; GABA, �-aminobutyric acid; GO, geneontology; P, postnatal day; RD, repeated doses; SD, single dose; SNR,signal to noise ratio; TX, hypothyroid rats.Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

0013-7227/08/$15.00/0 Endocrinology 149(8):3989–4000Printed in U.S.A. Copyright © 2008 by The Endocrine Society

doi: 10.1210/en.2008-0350

3989

at Inst Biomedicas Albertosols CSIC on March 9, 2010 endo.endojournals.orgDownloaded from

http://endo.endojournals.org

administration to hypothyroid rats on striatal gene expres-sion by microarray analysis. As a result, we identified novelgene targets of thyroid hormone. In terms of gene regulation,the effect of a single, large dose of T3 was more dramatic thanthat of multiple lower doses, probably indicating that am-plified gene responses to T3, as occurs in liver (12), are alsopresent in the adult brain. This observation is important toexplain the beneficial effects of large doses of thyroid hor-mone in bipolar disorders.

Materials and MethodsAnimals and treatment

Rats from the Wistar strain grown in our animal facilities were used.Protocols for animal handling were approved by the local institutionalAnimal Care Committee and followed the rules of the European Union.Animals were under temperature (22 � 2 C) and light (12-h light, 12-hdark cycle; lights on at 0700 h) controlled conditions and had free accessto food and water. Hypothyroidism was induced in adult male rats atpostnatal day (P) 50 by surgical thyroidectomy, taking care to spare theparathyroid glands (9). To ensure complete thyroidectomy, the rats weregiven the antithyroid compound 2-mercapto-1-methylimidazole (SigmaChemical Co., St. Louis, MO) 0.02% in the drinking water until the endof the experiment, on P75. Hypothyroid rats (referred to as TX) had lowserum levels of both T4 (0.34 � 0.13 ng/ml) and T3 (0.07 � 0.05 ng/ml),compared with normal rats of the same age (24.8 � 5.7 and 0.49 � 0.07ng/ml, respectively). Serum thyroid hormones were measured as de-scribed (13).

T3 (Sigma), dissolved in 0.05 m NaOH and diluted in saline con-taining 0.1% BSA, was administered ip to the TX rats according to twodifferent schedules: in one group T3 was given as a single dose of 25�g of T3 per 100 g body weight on P74, i.e. 24 h before the animalswere killed (SD group). Plasma T3 concentration 24 h after the in-jection was 4.35 � 1.93 ng/ml, corresponding to an estimated frac-tional occupancy of T3 nuclear receptors of 0.87, according to theformula: occupancy � plasma T3/(plasma T3 � 0.67) (14). The SDgroup was used to identify fast changes of gene expression takingplace after T3 administration to TX rats, under conditions of near fullsaturation of nuclear receptors for 24 h. A second group of TX ratswas also treated with T3 but for 5 d with single daily doses of 1.5 �gof T3 per 100 g body weight, starting on P70 (RD group). Plasma T3concentration 24 h after the last injection was 0.66 � 0.17 ng/ml andan estimated T3 receptor occupancy of 0.49. The effect of T3 treatmentwas monitored by Northern blotting analysis of liver type 1 deiodi-nase (D1) mRNA (15) and quantified by densitometry using theNational Institutes of Health Image J software (http://rsb.info.nih.gov/ij/), with correction for the housekeeping gene Gapdh. Bothtreatment schedules resulted in similar inductions of D1 mRNA (TX:0.40 � 0.05; SD: 1.30 � 0.20; RD: 1.15 � 0.20, P � 0.05 for T3 treatedvs. TX).

The animals, weighing about 200 g, were killed by decapitation underanesthesia with a mixture of ketamine and medetomidine (9) 24 h afterthe single T3 dose or the last of the multiple doses. The brain was rapidlyremoved and the striatum was isolated after separation from the internalcapsule and kept frozen at �80 C until RNA preparation.

RNA analysis

Total RNA was isolated individually from each animal using theTrizol procedure (Invitrogen, Carlsbad, CA), with an additional step of

chloroform extraction. The quality of RNA was analyzed using a Bio-Analyzer (Agilent, Santa Clara, CA). cDNA was prepared from 250 ngof RNA using the high-capacity cDNA reverse transcription kit (AppliedBiosystems, Foster City, CA). For quantitative PCR, a cDNA aliquotcorresponding to 5 ng of the starting RNA was used, with TaqmanAssay-on-Demand primers and the Taqman universal PCR master mix,No Amp Erase UNG (Applied Biosystems) on a 7900HT fast real-timePCR system (Applied Biosystems). The PCR program consisted in a hotstart of 95 C for 10 min, followed by 40 cycles of 15 sec at 95 C and 1 minat 60 C. PCRs were performed in triplicates, using the 18S gene asinternal standard and the 2-cycle threshold method for analysis (16).

Individual striatal RNA samples from six hypothyroid, four SD,and five RD rats were hybridized to separate Codelink microarrays(rat whole genome bioarray; GE Healthcare Europe GmbH, Munich, Ger-many). These arrays contain 33,849 probes, representing 14,519 uniquesequences. A limited survey was also performed using RNA pools fromfive animals of each condition and hybridized to the Rat ExpressionArray 230A (product 511036 from Affymetrix, Santa Clara, CA), con-taining 10,417 unique sequences. There were 9749 sequences common toboth platforms (see supplemental Fig. 1, published as supplemental dataon The Endocrine Society’s Journals Online web site at http://endo.endojournals.org). All procedures were as recommended by the man-ufacturers. Codelink hybridizations were performed at the Instituto deInvestigaciones Biomedicas, Madrid, whereas Affymetrix hybridiza-tions were performed at the University of Pisa.

Analysis of the data from microarray hybridizations

The data were analyzed using the R software (17) and packages fromthe Bioconductor project (http://www.bioconductor.org/) (18, 19). Thecodelink (20) and affy (21) packages were used for reading and prepro-cessing the arrays, genefilter (22) for data filtering and limma (23) forstatistical analysis. For Codelink arrays, raw intensities were exportedfor each array with the Codelink software and the files read into R.Background correction using the normexp method and quantile normal-ization was applied. Probes having a signal to noise ratio (SNR) below1 in all samples were removed from further analysis. Gene-wise inten-sities were fitted to a linear model with the experimental group (TX, RD,and SD) as factor and contrasts RD-TX and SD-TX were computed toidentify genes differentially expressed between the treatments and thereference group. Genes were selected as differentially expressed withPadjust � 0.05 (24). Data from the Affymetrix arrays were loaded into Rand processed using the robust multichip average (25) method. Geneswere selected based on fold change to compare the results with theCodelink data. For this task an absolute fold change of 1.6 was used. (0.7in log2 scale).

We performed an analysis of enriched gene ontology (GO) (26)categories to look for affected biological processes, molecular func-tions, or pathways using the PANTHER resource (27). A gene uni-verse was created by filtering from the list of all probes in the arraythose not expressed, i.e. with a SNR less than 1 in all samples. Afterthat, we translated the probe identifiers to Entrez Gene identifiers(removing any probe without Entrez Gene) and filtered those lackingGO annotations. Finally, we removed all duplicated Entrez Geneidentifications, obtaining 7,287 identifiers from the original 33,849probes (supplemental Table 1. For the test we translated the list ofdifferentially expressed genes to Entrez Gene and removed those notpresent in the universe.

FIG. 1. Effects of T3 on gene expression in the adult rat striatum. A, Two-dimensional representation of expression values (heat map) acrossall samples and differentially expressed genes. All intensities were normalized across the rows (subtracting the mean and dividing by the SD)to enable comparison between genes. The normalized log intensity values of the probes were centered to the median value of each probe setand colored on a range of �2 (red) and �2 (blue); yellow indicates intermediate value. Columns correspond to samples, whereas row correspondsto individual probe sets (in some cases several probe sets correspond to the same gene). Two main clusters, a and b, classify the differentiallyexpressed genes as having higher or lower expression after T3 treatment, respectively. The horizontal arrow shows a cluster of genes showinghigher expression after multiple doses of T3 (RD group) but not after a single dose (SD group). B, The profiles plot shown here contains thesame information found in the heat map but with a different perspective. This allows us to focus on the general tendency of selected gene clusters.Signal intensities are normalized as in the heat map and expression values and colored based on the overall pattern. For each pattern, a loess(locally weighted polynomial regression) fit line describes the overall profile of the corresponding group. This tendency profile is plotted

3990 Endocrinology, August 2008, 149(8):3989–4000 Diez et al. • Effect of T3 on Gene Expression in the Adult Striatum



SD candidate genes

RD candidate genes

A B

a

b

separately for genes up-regulated (red and green) and down-regulated (blue). The upper panel shows the expression values of each of the 222up-regulated sequences and 112 down-regulated sequences selected after the T3 single dose (SD group) in the individual arrays for the threeexperimental groups. The lower panel corresponds to the expression values of each of the 25 up-regulated sequences and the one that wasdown-regulated after the T3 repeated dose (RD group). In the RD group, the up-regulated genes follow two trends: one with little change inthe SD group (green line) and another that was also changed in the SD group (red line); there was only one gene decreased after RD (blue line).Most of the genes in cluster a of the heat map correspond to genes following the blue line in B. On the other hand, genes in cluster b in theheat map correspond to genes following the red line in B. Those genes indicated with the green line correspond to the cluster shown with anarrow in the heat map.

Diez et al. • Effect of T3 on Gene Expression in the Adult Striatum Endocrinology, August 2008, 149(8):3989–4000 3991



ResultsEffects of T3 on gene expression in the adult striatum

To analyze the effects of thyroid hormone on gene expres-sion in the adult rat striatum, T3 was administered to hypo-thyroid rats either acutely (SD) or as single daily doses (RD).RNA from individual rat striata was used to hybridizeCodelink microarrays. Both the SD and RD treatments re-sulted in changes of gene expression. Figure 1A shows thehierarchical clustering (heat map) of all the individual datarepresenting the differentially expressed sequences at a sig-nificance level of Padjust � 0.05. The complete, annotated heatmap is shown as supplemental Fig. 2. The columns containthe data from individual samples, which clustered in threegroups as a function of treatment. The rows show the relativeintensity of the probes, centered around the mean intensityfor all individual RNAs. The probes were grouped in twomain clusters: one group (cluster a) contained sequenceswhose expression increased after T3. The second group (clus-ter b) contained sequences whose expression was decreasedby T3. Visual inspection of the heat map also revealed that theeffect of the SD was more pronounced than the RD, although

in most cases there was also an effect but of lower intensityof the RD.

The comparisons between the effects of the SD vs. the RDare clearly seen in the plot profiles shown in Fig. 1B. Theseplots show the expression values along different arrays forthe probes selected as differentially expressed in the twotreatment groups. The signal intensities are scaled to thesame range so that they can be compared. The tendencyprofile [Loess curve (28)] is plotted separately for up-regu-lated (Fig. 1B, red and green) and down-regulated (Fig. 1B,blue) genes. The upper panel shows the behavior in each groupof the genes that changed after the single T3 dose (SD group).The lower panel shows the genes whose expression waschanged after the multiple T3 doses (RD group). The dataindicate that administration of a single high dose of T3 setsin motion large changes of gene expression, in contrast to themore discrete changes after treatment with the much lower,multiple doses.

Figure 2 shows the MA scatter plots (Fig. 2, upper panels)and the Venn diagram (Fig. 2, lower panel). The MA plotsrelate the log2 of the fold change (M) of the signal after T3

DRDSA

down-regulated

up-regulated

B

FIG. 2. Effects of single (SD) and multiple (RD) doses of T3 on gene expression after administration to TX animals. A, MA plots, relating thelog2 of the fold change (M, in ordinate) vs. the mean of log intensity of the signal (A, in abscissa) of SD vs. TX and RD vs. TX. Red and orangedots represent probe sets having SNRs of 1 or below. Black dots are probe sets with SNR above 1. The blue dots represent the differentiallyexpressed sequences, with Padjust � 0.05. B, Venn diagram showing the quantitative relation between the number of sequences differentiallyexpressed in the SD and RD groups.




treatment with its mean log-expression level (A) in thetreated and untreated groups. The dotted horizontal lines rep-resent a fold change of 2 for up-regulated genes or 50% fordown-regulated genes. The blue dots represent the differen-tially expressed probe sets with Padjust � 0.05. In the SDgroup, 334 sequences were differentially expressed with re-spect to the hypothyroid animals. From them, 222 sequenceswere up-regulated and 112 sequences down-regulated. Aftersubtracting 80 nonannotated sequences and 17 repeatedprobes, there were 149 up-regulated and 88 down-regulatedgenes. Many of the statistically significant, differentially ex-pressed genes had absolute M values as low as 0.5, indicatingmoderate changes of gene expression. After filtering the dataon the basis of fold change, there were 91 up-regulated genesand 53 down-regulated genes with absolute M values greaterthan 0.7.

The RD caused changes of expression of 26 sequences.After subtracting the nonannotated and repeated sequences,18 genes were up-regulated and one was down-regulated[Gja7, or connexin 45 (29)]. As shown in the intersectionbetween the SD and RD in the Venn diagram, 17 sequenceswere up-regulated with both treatment schedules. This set of17 sequences corresponded to 10 unique genes: the canna-binoid receptor Cnr1, an important G protein-coupled re-ceptor (30), recently linked to hyperactivity of neonatal hy-pothyroid rats (31); angiotensin converting enzyme, Ace, acomponent of brain renin-angiotensin system (32); the sonichedgehog transcriptional effector Gli1, linked to neuropro-tection in the adult brain (33); the sonic hedgehog responsivegene Ith3 [inter-� trypsin inhibitor (34)]; the extraneuronalmonoamine transporter Slc22a3, implicated in several neu-rological disorders (35); the voltage gated sodium channel�4-subunit, Scn4b, involved in electrical signaling and celladhesion and defective in Huntington’s disease (36, 37); thesmall heat shock protein Hspb6 (38); Grifin, or galactin-relatedinterfiber protein originally described in the lens (39); andLoc290651, similar to myoinositol 1-phosphate synthase A1,and RGD1305038_predicted, similar to serum-induciblekinase.

The mean expression of the eight sequences specificallyincreased by RD with P � 0.05 is represented as a green linein the plot profile of Fig. 1B. Seven of these sequences couldbe observed as a prominent cluster in the heat map (Fig. 1,horizontal arrow). Three of these genes are related to red bloodcells (Loc287167, Hba1, and AlaS2); Ceacam1 encodes a celladhesion molecule (40); Prph1 encodes the intermediate fil-ament and Akt substrate peripherin (41, 42); and Cd52 en-codes a surface antigen.

The complete set of genes that were differentially ex-pressed in the SD and RD groups, selected on the basis ofstatistical significance is shown as supplemental Table 2.From these data, genes having a fold change of at least 1.6 inany of the two treatment groups are displayed in Table 1.This table shows the fold change in the SD and the RD groupsand the A value. The A value corresponds to the mean of thelog2 intensity values over all samples and is a measure of theaverage expression level. Values with Padjust � 0.05 are high-lighted. The genes were categorized according to the bio-logical processes or molecular function assigned by GOannotation (26). The GO annotation suggested that the dif-

ferentially expressed genes were involved in a variety ofbiological processes.

To determine the statistical significance of these observa-tions, we used PANTHER analysis to identify those catego-ries significantly over- or downrepresented within the groupof differentially expressed genes in relation to the universeof expressed genes in the striatum as found in the arrays(7287; see Materials and Methods). The analysis (supplementalTable 3) showed the pathway, biological process and mo-lecular function categories with over- or downrepresentedcandidate genes with P � 0.05. Within the pathway category,we found three genes related to the circadian clock system(Per1, Per2, Nr1d2); heterotrimeric G protein signalingpathways (Gq� and Go�-mediated pathways: Rgs2, Rgs14,Rgs9, Rasgrp1, Arhgef3_predicted, Rhoc_predicted, Rap1ga1,Rasgrp2_predicted); oxidative stress response (Dusp1,Dusp5, Map2k3, Dupd1_predicted); phenylethylamine deg-radation (Aldh1a1, Doxl1); and MAPK pathway (Map2k3,Fos, Rps6ka4_predicted). Within the biological process andmolecular function categories, we found, among others,genes involved in regulation of phosphate metabolism,signal transduction, tyrosine kinase signaling pathways,extracellular transport and import, cell structure, and neu-ronal activity.

Confirmation of microarray data

We used two approaches to confirm the data. First, wechecked the reproducibility of the assay in a different plat-form. Reproducibility of microarray data across differentplatforms and laboratories is dramatically dependent on thecriteria used to select the differentially expressed genes (43).In our studies we analyzed individual RNAs and applied thelimma statistics to account for variability of the data andtherefore selected the candidate genes on the basis of ad-justed P value. On the other hand, it has been noted thatselection on the basis on fold change achieves better repro-ducibility among different platforms and biological groupsthan a selection based on t statistics (44, 45).

Therefore, as a secondary screening for partial validationof the data and to select sequences for further confirmationby PCR, we performed an independent analysis using theAffymetrix platform. In this assay, we used pools of RNAinstead of individual samples. Striatal RNA from five ani-mals of each group were pooled and hybridized to the Af-fymetrix rat expression arrays 230A. Candidate genes wereselected on the basis of an increase or decrease of at least1.6-fold and compared with the candidate genes previouslyselected from the Codelink arrays. These comparisons werelimited by the fact that some genes were uniquely present inone of the two different arrays (see supplemental Fig. 1 andsupplemental Table 4). Fifty two differentially expressedgenes from the Codelink platform were absent from theAffymetrix arrays, and 20 genes with an M value above 0.7in the Affymetrix arrays were absent from the Codelinkarrays. Among the latter, Nrgn and Tubb3, known to bedependent of the thyroid status (11, 46), were found as in-duced by SD and RD treatments in the Affymetrix arrays (notshown) but were not present as probes in the Codelinkarrays.




TABLE 1. Differentially expressed genes grouped by biological process (BP; GO)

Entrezid Symbol Name SD RD A

BP: regulation of transcription, DNA dependent140589 Gli1 GLI-Kruppel family member GLI1 3.43 2.35 4.45306873 Rreb1_predicted Ras responsive element binding protein 1 (predicted) 2.91 1.13 5.6224309 Dbp D site albumin promoter binding protein 2.66 �1.07 5.9160563 hr Hairless homolog (mouse) 2.64 1.00 6.02298894 Mycn V-myc myelocytomatosis viral related oncogene, neuroblastoma derived

(avian) (mapped)2.30 1.17 5.44

313575 Heyl_predicted Hairy/enhancer-of-split related with YRPW motif-like (predicted) 2.23 1.49 4.8924309 Dbp D site albumin promoter binding protein 2.00 1.13 9.47361715 Rps6ka4_predicted Ribosomal protein S6 kinase, polypeptide 4 (predicted) 1.93 1.25 5.91117560 Klf9 Kruppel-like factor 9 1.91 1.55 8.67117560 Klf9 Kruppel-like factor 9 1.83 1.48 10.58308435 Cic_predicted Capicua homolog (Drosophila) (predicted) 1.67 1.20 8.21497984 RGD1566329 predicted Similar to zinc finger protein 652 (predicted) �1.13 6.73 5.13287422 Per1 Period homolog 1 (Drosophila) �1.83 �1.23 6.6124330 Egr1 Early growth response 1 �1.99 �1.02 10.525129 Egr4 Early growth response 4 �2.16 1.13 8.31114519 Nfil3 Nuclear factor, IL-3 regulated �2.27 �1.06 6.01304741 Tcfcp2l1_predicted Transcription factor CP2-like 1 (predicted) �2.35 �1.13 2.81314322 Fos FBJ murine osteosarcoma viral oncogene homolog �3.07 �1.16 7.6679240 Nr4a1 Nuclear receptor subfamily 4, group A, member 1 �3.71 �1.16 8.4658853 Nr4a3 Nuclear receptor subfamily 4, group A, member 3 �3.81 �1.12 5.5758853 Nr4a3 Nuclear receptor subfamily 4, group A, member 3 �4.29 �1.03 7.26114090 Egr2 Early growth response 2 �4.96 �1.07 6.5129546 Homer1 Homer homolog 1 (Drosophila) �5.58 �1.30 6.64

BP: intracellular signaling cascade64032 Ctgf Connective tissue growth factor 2.99 1.64 5.360669 Cmklr1 Chemokine-like receptor 1 2.31 1.22 5.8125248 Cnr1 Cannabinoid receptor 1 (brain) 2.11 2.03 8.55361715 Rps6ka4_predicted Ribosomal protein S6 kinase, polypeptide 4 (predicted) 1.93 1.25 5.91311118 Tlk1_predicted Tousled-like kinase 1 (predicted) �1.64 �1.11 4.91114108 Pdlim3 PDZ and LIM domain 3 �1.80 1.11 6.8924716 Ret Ret protooncogene �1.87 �1.60 4.1229546 Homer1 Homer homolog 1 (Drosophila) �5.58 �1.30 6.64

Small GTPase-mediated signal transduction306873 Rreb1_predicted Ras responsive element binding protein 1 (predicted) 2.91 1.13 5.62310838 Bcar3_predicted Breast cancer antiestrogen resistance 3 (predicted) 2.57 1.39 6.61310838 Bcar3_predicted Breast cancer antiestrogen resistance 3 (predicted) 2.39 1.33 6.61114203 Aps Adaptor protein with pleckstrin homology and src homology 2 domains 2.31 1.12 5.39303653 Cdc42ep4_predicted CDC42 effector protein (� GTPase binding) 4 (predicted) 2.25 1.19 6.77303653 Cdc42ep4_predicted CDC42 effector protein (� GTPase binding) 4 (predicted) 2.23 1.29 6.84171099 Rasd2 RASD family, member 2 1.88 1.45 10.03299824 RGD1305255 predicted Similar to CG3996-PA (predicted) 1.79 1.53 5.61360571 Rab34 RAB34, member of RAS oncogene family 1.67 1.15 6.57192126 Dab2ip Disabled homolog 2 (Drosophila) interacting protein 1.65 1.15 8.78292750 Plekhg2_predicted Pleckstrin homology domain containing, family G (with RhoGef domain)

member 2 (predicted)�1.66 1.04 6.63

308821 Rab30 RAB30, member RAS oncogene family �2.81 �1.24 3.71686142 LOC686142 Similar to Ral guanine nucleotide dissociation stimulator (RalGEF)

(RalGDS)�2.85 1.13 6.01

G protein-coupled receptor protein signaling pathway29481 Rgs9 Regulator of G-protein signaling 9 2.75 1.62 5.2629481 Rgs9 Regulator of G-protein signaling 9 2.71 1.78 6.0184583 Rgs2 Regulator of G-protein signaling 2 �1.66 1.09 8.9

Protein tyrosine phosphatase activity246781 Ptpn7 Protein tyrosine phosphatase, nonreceptor type 7 �1.64 �1.04 6.65171109 Dusp5 Dual specificity phosphatase 5 �2.03 �1.06 5.02295338 Ptpn22_predicted Protein tyrosine phosphatase, nonreceptor type 22 (lymphoid) (predicted) �1.65 1.05 4.73114856 Dusp1 Dual-specificity phosphatase 1 �2.06 �1.39 7.01

BP: rhythmic process24309 Dbp D site albumin promoter binding protein 2.66 �1.07 5.9124309 Dbp D site albumin promoter binding protein 2.00 1.13 9.47287422 Per1 Period homolog 1 (Drosophila) �1.83 �1.23 6.6129461 Vgf VGF nerve growth factor inducible �1.95 �1.08 9.05114090 Egr2 Early growth response 2 �4.96 �1.07 6.51

BP: cation transport29504 Slc22a3 Solute carrier family 22, member 3 2.73 2.35 5.2683500 Slc22a8 Solute carrier family 22 (organic anion transporter), member 8 2.69 1.72 4.28

(Continues)




TABLE 1. Cont.


29726 Slc22a5 Solute carrier family 22 (organic cation transporter), member 5 2.16 1.02 4.96315611 Scn4b Sodium channel, voltage-gated, type IV, � 1.97 1.56 10.65315611 Scn4b Sodium channel, voltage-gated, type IV, � 1.88 1.65 9.67170546 Ryr3 Ryanodine receptor 3 1.84 1.09 7.56170546 Ryr3 Ryanodine receptor 3 1.83 1.19 6.38252919 Slc38a3 Solute carrier family 38, member 3 1.82 1.13 5.33170924 Abcc4 ATP-binding cassette, subfamily C (CFTR/MRP), member 4 1.67 1.15 4.91155205 Slc36a1 Solute carrier family 36 (proton/amino acid symporter) member 1 �2.16 1.01 3.7385267 Slc24a3 Solute carrier family 24 (sodium/potassium/calcium exchanger), member 3 �2.41 1.00 3.29

BP: cell organization and biogenesis315131 Kdelr3_predicted KDEL (Lys-Asp-Glu-Leu) endoplasmic reticulum protein retention

receptor 3 (predicted)6.50 1.52 3.01

63878 Stc2 Stanniocalcin 2 5.54 3.41 2.7325602 Tnxa Tenascin XA 5.21 1.93 5.97316638 Sned1 Insulin responsive sequence D binding protein-1 2.93 1.67 5.76362650 Fblim1 Filamin binding LIM protein 1 2.81 1.96 3.45171112 Wfdc1 WAP four-disulfide core domain 1 2.17 1.27 7.6929317 Csrp2 Cysteine and glycine-rich protein 2 2.04 1.40 7.1689806 Tie1 Tyrosine kinase with immunoglobulin-like and EGF-like domains 1 1.91 1.14 6.48307235 RGD1311910 predicted Similar to hypothetical p38 protein (predicted) 1.88 �1.08 6.5285251 Col18a1 Procollagen, type XVIII, �1 1.80 1.13 4.39308345 Suv420h2_predicted Suppressor of variegation 4–20 homolog 2 (Drosophila) (predicted) 1.68 1.22 7.43360537 Tom1l2_predicted Target of myb1-like 2 (chicken) (predicted) 1.68 1.01 4.85362584 Col9a2_predicted Procollagen, type IX, �2 (predicted) 1.68 1.13 5.9481613 Ceacam1 CEA-related cell adhesion molecule 1 1.46 2.53 4.71266706 Gja7 Gap junction membrane channel protein �7 �1.09 �1.78 4.1124688 Prph1 Peripherin 1 �1.10 1.92 8.11362873 Plxnc1_predicted Plexin C1 (predicted) �1.62 �1.16 5.11287925 Pkp2 Plakophilin 2 �1.62 1.05 5.61306081 Pcdh20_predicted Protocadherin 20 (predicted) �1.65 �1.03 7.2754323 Arc Activity regulated cytoskeletal-associated protein �2.00 1.15 7.86306548 Hook3 Hook homolog 3 (Drosophila) �2.01 �1.18 2.8679110 Mepe Matrix extracellular phosphoglycoprotein with ASARM motif (bone) �2.58 �1.11 2.99

BP: cellular lipid metabolism89826 Rpe65 Retinal pigment epithelium 65 3.23 1.61 3.3689826 Rpe65 Retinal pigment epithelium 65 2.16 1.71 4.0884575 Fads1 Fatty acid desaturase 1 1.93 1.47 6.81117243 Acsl6 Acyl-CoA synthetase long-chain family member 6 1.82 1.36 8.9594338 Smpd3 Sphingomyelin phosphodiesterase 3, neutral 1.77 1.28 8.76170465 Acaa2 Acetyl-coenzyme A acyltransferase 2 (mitochondrial 3-oxoacyl-coenzyme

A thiolase)1.67 �1.09 7.39

25702 Pnlip Pancreatic lipase 1.46 1.91 7.32304741 Tcfcp2l1_predicted Transcription factor CP2-like 1 (predicted) �2.35 �1.13 2.81

BP: protein metabolism24310 Ace Angiotensin I-converting enzyme (peptidyl-dipeptidase A) 1 4.66 2.79 6.87295394 Dph5 DPH5 homolog (S. cerevisiae) 2.20 1.35 2.8425697 Ctsl Cathepsin L 1.72 1.21 5.91364948 Asphd2 Aspartate �-hydroxylase domain containing 2 1.69 1.25 7.67308345 Suv420h2_predicted Suppressor of variegation 4–20 homolog 2 (Drosophila) (predicted) 1.68 1.22 7.43300981 Acy1 Aminoacylase 1 1.67 1.28 7.66310467 Tiparp_predicted TCDD-inducible poly(ADP-ribose) polymerase (predicted) �1.85 �1.16 5.91303287 Fbxo39 F-box protein 39 �2.27 1.20 4.2629154 Capn2 Calpain 2 �2.39 �1.11 6.91

BP: others303772 Slc16a6 Solute carrier family 16 (monocarboxylic acid transporters), member 6 5.78 2.11 6.89303772 Slc16a6 Solute carrier family 16 (monocarboxylic acid transporters), member 6 4.86 2.25 4.37303772 Slc16a6 Solute carrier family 16 (monocarboxylic acid transporters), member 6 4.29 1.21 2.89293628 RGD1308955 Similar to Cc1–9 3.81 2.04 3.31314627 RGD1305038 predicted Similar to serine/threonine-protein kinase SNK (serum inducible kinase)

(predicted)3.25 1.71 6.62

296048 Agpat7_predicted 1-Acylglycerol-3-phosphate O-acyltransferase 7 (lysophosphatidic acidacyltransferase, �) (predicted)

2.51 1.71 7.01

192245 Hspb6 Heat shock protein, �-crystaline-related, B6 2.22 1.60 7.7250693 Itih3 Inter-� trypsin inhibitor, heavy chain 3 1.95 2.53 8.61287437 Tnfsf13 TNF (ligand) superfamily, member 13 1.91 1.17 6.31303601 Cyb561_predicted Cytochrome b-561 (predicted) 1.88 1.39 8.254283 Pfkfb4 6-Phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 1.83 1.12 6.1124188 Aldh1a1 Aldehyde dehydrogenase family 1, member A1 1.67 1.55 5.33

(Continues)




TABLE 1. Cont.


260327 Fndc5 Fibronectin type III domain containing 5 1.67 1.29 7.93293587 RGD1311186 predicted Similar to RIKEN cD 1810014F10 gene (predicted) 1.66 1.30 7.16303242 DLP2 Dynein-like protein 2 1.62 1.13 6.1125748 Alas2 Aminolevulinic acid synthase 2 1.12 5.06 4.27287167 LOC287167 Globin, � �1.04 11.39 7.2225632 Hba-a1 Hemoglobin-�, adult chain 1 �1.39 7.46 4.46304017 Tomm70a Translocase of outer mitochondrial membrane 70 homolog A (yeast) �1.66 �1.02 10.425741 Pfkl Phosphofructokinase, liver, B-type �1.97 �1.02 5.91171128 Gchfr GTP cyclohydrolase I feedback regulator �2.46 �1.75 3.93292686 Fbxo46 F-box protein 46 �2.77 1.25 5.6729625 Crhbp Corticotropin releasing hormone binding protein �3.16 �1.22 5.21

MF: others301073 Gup1_predicted Gup1, glycerol uptake/transporter homolog (yeast) (predicted) 4.63 1.71 5.73309208 RGD1564983 predicted Similar to leucine rich repeat containing 10 (predicted) 3.86 2.00 8.01500590 LOC500590 Similar to T cell antigen 4–1BB precursor-mouse 3.56 1.16 4.73117130 Grifin Galectin-related interfiber protein 2.41 5.31 3.27363931 Gtpbp6_predicted GTP binding protein 6 (putative) (predicted) 2.19 1.26 9.32288689 RGD1566317 predicted Similar to tescalcin (predicted) 2.04 1.44 9.1363886 Abcb9 ATP-binding cassette, subfamily B (MDR/TAP), member 9 1.93 1.45 6.77367747 RGD1564763 predicted Similar to nudix (nucleoside diphosphate linked moiety X)-type motif 11

(predicted)1.85 1.38 5.85

361348 RGD1559803 predicted Similar to maestro (predicted) 1.73 1.27 6.59171451 Mosc2 MOCO sulphurase C-terminal domain containing 2 1.69 1.07 8.92362597 RGD1308876 predicted Similar to 2610027C15Rik protein (predicted) 1.67 1.07 5.11361552 Wtip_predicted WT1-interacting protein (predicted) 1.65 �1.04 6.5789789 Lbr Lamin B receptor �1.64 �1.15 4.65306860 Gcnt2 Glucosaminyl (N-acetyl) transferase 2, I-branching enzyme �1.69 �1.05 7.2864629 Porf1 Preoptic regulatory factor 1 �1.85 �1.65 7.41312316 Doxl1 Diamine oxidase-like protein 1 �2.53 1.22 4.48366624 Cfl2_predicted Cofilin 2, muscle (predicted) �2.83 1.13 2.71360949 Cpeb2_predicted Cytoplasmic polyadenylation element binding protein 2 (predicted) �3.61 1.04 4.48

No data307494 RGD1305640 predicted Similar to RIKEN cD 2410080P20 (predicted) 2.79 1.80 3.61289235 Slamf9_predicted SLAM family member 9 (predicted) 2.68 1.41 4.31308794 RGD1310371 Similar to RIKEN cD 1700026D08 2.60 1.66 6.11300317 RGD1311154 predicted Similar to hypothetical protein FLJ12242 (predicted) 2.35 1.49 9.42361267 Akr1cl1_predicted Aldo-keto reductase family 1, member C-like 1 (predicted) 2.33 1.27 4.69361853 RGD1311294 predicted Similar to hypothetical protein C6orf60 (predicted) 2.30 1.48 5.59691504 LOC691504 Similar to Zinc finger protein ZFPM1 (Zinc finger protein multitype 1)

(friend of GATA protein 1) (friend of GATA-1) (FOG-1)2.27 1.33 6.99

500200 RGD1561512 predicted Similar to bHLH factor Math6 (predicted) 2.25 1.49 3.89313837 RGD1562284 predicted Similar to glutaminyl-peptide cyclotransferase precursor (QC)

(predicted)2.13 1.67 6.75

499745 LOC499745 Similar to Notch-regulated ankyrin repeat protein 1.96 1.20 4.64266764 Tbkbp1 TBK1 binding protein 1 1.89 1.42 6.62361519 RGD1564862 predicted Similar to hypothetical protein MGC51082 (predicted) 1.84 1.27 8.61303671 Fads6_predicted Fatty acid desaturase domain family, member 6 (predicted) 1.72 1.18 7.71362246 Trp53inp2 Tumor protein p53 inducible nuclear protein 2 1.68 1.39 9.41500059 RGD1559885 predicted Similar to hypothetical gene supported by BC063892 (predicted) 1.65 1.09 6.93117054 Cd52 CD52 antigen 1.59 6.54 2.76691478 LOC691478 Similar to copine-6 (copine VI) (Neuronal-copine) (N-copine) �1.71 �1.39 8.28305549 Peli1 Pellino homolog 1 (Drosophila) �1.71 1.00 7.21499839 RGD1564664 predicted Similar to LOC387763 protein (predicted) �1.73 �1.17 9.48494538 LOC494538 ABP� �1.78 �1.24 5.12691164 LOC691164 Similar to pellino homolog 1 (Drosophila) �1.92 �1.04 7.27363496 Armcx6 Armadillo repeat containing, X-linked 6 �1.96 �1.13 6.91500204 RGD1562515 predicted Similar to RIKEN cD 4931417E11 (predicted) �1.97 �1.32 4.32361003 Dupd1_predicted Dualspecificity phosphatase and proisomerase domain containing 1

(predicted)�2.28 1.01 2.91

298605 RGD1566169 predicted Similar to hypothetical protein MGC37938 (predicted) �2.50 �1.60 5.62

Shown are the genes with a fold change above 1.6 or below �1.6, in any of the SD or RD groups. Fold changes are represented in naturalscale, computing the magnitude with the most abundant transcript in the numerator. The sign indicates the direction of the change:positive values refer to greater transcript abundance in RD or SD respect to the reference (TX), whereas negative values indicate lessabundance. Shown also is the A value: mean of the log2 intensity values over all samples as a measure of the average expression level.The genes are grouped by the BP categories of GO. Repeated probes were not omitted. In bold, the fold change of probes with Padjust �0.05. MF, Molecular functions.




From the genes common to both platforms (supplementalTable 5), five were up-regulated by both T3 treatment sched-ules (see below), and three were up-regulated by the RD only(Hba1, Alas2, Loc287167). From the SD treatments, 26 up-regulated and 14 down-regulated genes were confirmed inboth platforms. Among the SD up-regulated genes, hr (orhairless), Rasd2 (also known as Rhes), and Klf9 (also knownas BTEB) are already known to be thyroid hormone-regu-lated genes. Among the SD down-regulated genes, therewere several early response genes (Egr1, Arc, Dusp1, Egr4,Fos, Nr4a1, Nr4a3, Egr2, and Homer1).

For biological confirmation using Taqman PCR, we useda different group of rats from the ones used for the arrays.The genes selected for confirmation included the five genesthat were up-regulated by both SD and RD in both platforms:Ace (angiotensin converting enzyme), Hspb6 (heat shock pro-tein B6), Cnr1 (cannabinoid receptor 1), Itih3 (inter-� trypsininhibitor), and Scn4b (voltage-gated sodium channel IV). Wealso checked Rgs9 (regulator of G protein signaling 9), astriatum-specific gene (47), which shows a robust increase bySD; Klf9 (Kruppel-like factor, also known as basic transcrip-tion factor binding protein, or BTEB), a gene previouslyshown to be regulated by thyroid hormone in several celltypes (48); Dbp (D-site binding protein); Rasgrp1 (also knownas CalDAG-GEF1); and the early response genes Nr4a1(alsoknown as NGFIB), Arc, Dusp1, Egr1 (also known as NGFIA),and Homer1. The results, shown in Fig. 3, show that there wasa good correlation between the results from the arrays andthe PCR, both in the direction and magnitude of expressionchanges.

Discussion

In this article we provide, for the first time, a comprehen-sive list of genes that are differentially expressed by thyroid

hormone in the adult striatum. Our approach involved theanalysis of gene expression after administration of T3 tohypothyroid rats.

Most previous studies on the effects of T3 on brain geneexpression have been limited to the postnatal period, and fewgenes were known to be sensitive to T3 in the adult brain. Inone study, Haas et al. (49) used a limited set of 1224 neural-specific genes and found that hyperthyroidism induced onlymodest changes in the expression of 11 genes. This study isnot comparable with ours, first because of the limited num-ber of genes in the array, second because total brain wasanalyzed, and third because T3 was given at doses 10-foldhigher than in this work and for 10 consecutive days so thatthe animals were hyperthyroid.

One of the earliest genes identified as regulated by T3 inthe adult brain is Nrgn, a gene confirmed to be regulated atthe transcriptional level in cultured neurons (50) and dem-onstrated to be a sensitive marker of T3 action. Nrgn was notrecovered initially in the list of candidate genes because itwas not present as a hybridizable probe in the arrays. How-ever, it was present in the Affymetrix chips and was detectedafter the acute and multiple-dose treatments, in agreementwith previous studies. Other genes present as candidategenes and previously shown to be regulated by thyroid hor-mones in other paradigms include Klf9 (48), Egr1 (51), themitochondrial importer Tomm70 (52), the transcription factorHr (53), the small G protein Rasd2 (also known as Rhes) (54),the transcription factor Nr4a1 (55), the cannabinoid receptorCnr1 (31), and the dual-specificity phosphatase Dusp1 (56).

Early studies by Oppenheimer and colleagues (57) showedthat the response of some genes, such as pituitary GH, fol-lowed a linear response in relation to nuclear occupancy afterT3 administration. Expression of these genes did not increaseabove the baseline euthyroid level. For other genes, exem-

-8.00

-6.00

-4.00

-2.00

0.00

2.00

4.00

6.00

Fo

ld C

han

ge

Codelink

Affymetrix

RT-PCR

0.00

1.00

2.00

3.00

4.00

5.00

6.00

Fo

ld C

han

ge

Nr4

a1

Arc

Du

sp1

Eg

r1

Ho

mer

1

Ras

grp

1

Db

p

Cn

r1

Klf

9

Scn

b4

Ace

Hsp

b6

Rg

s9

Itih

3

Cn

r1

Klf

9

Scn

b4

Ace

Hsp

b6

Rg

s9

Itih

3

SD

RD

FIG. 3. Correlation between data from Codelink andAffymetrix hybridizations and Taqman PCR assays ofselected differentially expressed genes from the SD andRD groups.




plified by liver enzymes, the response was nonlinear andamplified with further increases of receptor occupancy abovethe euthyroid level of about 50% (12).

Based on these studies, the dosage and timing of T3 ad-ministration was chosen by us so that two different situationswere reached. The single dose was intended to result in nearfull saturation of nuclear receptors for 24 h. Twenty-fourhours after injection, the fractional occupancy of nuclearreceptors was calculated to be 0.86. In the case of multipleinjections, the fractional occupancy 24 h after the last injec-tion was 0.49, which is the physiological occupancy of T3receptors in liver (12). The single injection protocol shouldlikely select for fast, linear, and amplified responses after T3,whereas the multiple injection dosage should identify linear,steady-state responses. Therefore, it is not surprising that thenumber of candidate genes was larger after the acute injec-tion of T3 than after the multiple daily doses. In the RD group,there were one down-regulated gene and eight up-regulatedgenes. The only down-regulated gene, Gja7, encodes a gapjunction membrane channel protein expressed in several re-gions of the central nervous system but not in the striatum(58). This agrees with the almost no expression found in theRD-treated rats, indicating that thyroid hormone is involvedin maintaining Gja7 repression under normal conditions.Three of the up-regulated genes were of red blood cell origin.

The significance of this finding is uncertain. Interestingly,Hba1 has also been found previously to be regulated by T3

in the cerebellum in vivo and in primary cultures of cerebellargranular cells, presumably at the transcriptional level (56).The functional significance of the regulation of the rest ofgenes comprising this small set of candidate genes is un-certain because they do not share common functionalproperties.

Despite the above, the expression of many genes after theRD falls between the levels of the hypothyroid and acutelytreated rats, indicating that the group of genes sensitive toeach of the treatment schedules is qualitatively similar. Ac-tually, we calculate that 52% of the differentially expressedgenes in the SD were changed in the RD by at least 20%, andin the same direction. This justifies the functional analysisperformed using the PANTHER resource from the list of allcandidate genes found under the two treatment schedules.On the other hand, the overlap between the SD and RDsuggests that the genes induced in both treatment schedulesare regulated at the transcriptional level.

PANTHER analysis of the significance of representedpathways disclosed three genes of the circadian clock system.Although not present in the PANTHER analysis, anotherregulated gene, Dbp, has been described as a circadian gene(59). Also, other genes related to the wakefulness (Fos, Arc,Hsps, Egr1, Homer) (59) were sensitive to T3. Because at leastone of these genes, Dbp, has been proposed as a candidategene in bipolar disorders (60), it is tempting to speculate thatthe beneficial effects of large doses of thyroid hormones inbipolar disorders is mediated, at least in part, through reg-ulation of the expression of these genes.

Another group of candidate genes have important roles inthe physiology of striatal neurons. There are 314 genes ofenriched expression in the striatum (61), from which 21 weredifferentially expressed after T3 treatment (supplemental Ta-ble 6). Some of these genes are defective in Huntington’sdisease (37), such as Scn4b, Rasd2, Rasgrp1, Klf9, and Rgs9. Thelatter is also one of the 10 most enriched genes in the striatumin relation to other brain regions (61). Central to striatalneuron signaling is the regulation of dopamine and cAMP-regulated phosphoprotein (DARPP)-32 phosphorylation[Fig 4 (62)]. Phospho-DARPP-32 levels are controlled by thecAMP-protein kinase A pathway, which promotes phos-phorylation, and by the Ca2�-calcineurin pathway, whichpromotes dephosphorylation. Rgs9 and Rasd2 are involved inG protein signaling regulating the cAMP pathway (47, 63),whereas Nrgn is involved in the regulation of the Ca2�/calmodulin pathway (64). In addition, sodium channels andsome of the early response genes, such as Arc, Homer, andDusp1, are also regulated by phospho-DARPP-32. A largefraction of the candidate genes found in the screening are alsoinvolved in intracellular signaling cascades involving G pro-tein and cation transport. Therefore, thyroid hormone is in-volved in maintaining an optimal signaling in the striatumby acting at multiple control points.

In conclusion, we have identified for the first time a set ofgenes whose expression is dependent on thyroid hormone inthe adult striatum. We believe that this work opens the wayfor more detailed analysis of the effects of thyroid hormone

MAPK

NMDA

PKA

cAMP

D2

PP-2BDARPP-32

P-DARPP-32

PP-1P-CREBEarly genes

D1

Ion channels and receptors

Ca2+

2

1

CaM

4

5

VDCC

PLC

CaMKIV

3

FIG. 4. Role of thyroid hormone in the control of signaling in thestriatum. Shown is a simplified scheme of signaling pathways in thestriatum. DARPP-32 phosphorylation is under control of the cAMPpathway, which promotes phosphorylation, and by the Ca2�-calmod-ulin (CaM) pathway, which promotes dephosphorylation. cAMP pro-duction is under control of G protein-coupled membrane receptorssuch as dopamine receptors (D1, D2) and others not included in thefigure, such as opiate receptors, adenosine, and serotonin receptors.Voltage-dependent Ca2� channels (VDCC), N-methyl-D-aspartate(NMDA) glutamate receptors, dopamine 2 receptors (D2), and GABAAreceptors (not shown) regulate intracellular Ca2� concentration andCaM activation. Genes regulated by thyroid hormone are involved indifferent control points of the signaling cascade, as shown by thearabic numerals: 1, G protein-coupled receptors; 2, Ca2� signalingand CaM activation; 3, MAPK pathways; 4, early gene transcription;5, ion channels. PLC, Phospholipase C; PKA, protein kinase A; PP-2B,calcium and calmodulin-dependent protein phosphatase, or cal-cineurin; CaMKIV, calcium and calmodulin-dependent kinase IV;PP-1, protein phosphatase 1.




in the adult brain and their beneficial therapeutic effects inaffective disorders.

Acknowledgments

The technical expertise of Eulalia Moreno and Ana Torrecilla is grate-fully acknowledged.

Received March 13, 2008. Accepted April 30, 2008.Address all correspondence and requests for reprints to: Beatriz

Morte and Juan Bernal, Instituto de Investigaciones Biomedicas,Arturo Duperier 4, 28029 Madrid, Spain. E-mail: [email protected];or [email protected].

This work was supported by Grant BFU2005-01740 from the Ministryof Education and Science of Spain, the European Union IntegratedProject CRESCENDO (Consortium for Research on Nuclear Receptors inDevelopment and Aging) Grant LSHM-CT-2005-018652, and the Centrode Investigacion Biomedica en Red de Enfermedades Raras, Instituto deSalud Carlos III.

Present address for D.D.: Bioinformatics Center, Institute for Chem-ical Research, Kyoto University, Uji, Kyoto 6110011, Japan.

Disclosure Statement: The authors have nothing to disclose.

References

1. Oppenheimer JH, Schwartz HL 1997 Molecular basis of thyroid hormone-dependent brain development. Endocr Rev 18:462–475

2. Bernal J 2005 Thyroid hormones and brain development. Vitam Horm 71:95–122

3. Bernal J 2007 Thyroid hormone receptors in brain development and function.Nat Clin Pract Endocrinol Metab 3:249–259

4. Simon NM, Blacker D, Korbly NB, Sharma SG, Worthington JJ, Otto MW,Pollack MH 2002 Hypothyroidism and hyperthyroidism in anxiety disordersrevisited: new data and literature review. J Affect Disord 69:209–217

5. Baumgartner A, Bauer M, Hellweg R 1994 Treatment of intractable non-rapidcycling bipolar affective disorder with high-dose thyroxine: an open clinicaltrial. Neuropsychopharmacology 10:183–189

6. Bauer M, London ED, Rasgon N, Berman SM, Frye MA, Altshuler LL,Mandelkern MA, Bramen J, Voytek B, Woods R, Mazziotta JC, Whybrow PC2005 Supraphysiological doses of levothyroxine alter regional cerebral me-tabolism and improve mood in bipolar depression. Mol Psychiatry 10:456–469

7. Bauer M, Heinz A, Whybrow PC 2002 Thyroid hormones, serotonin andmood: of synergy and significance in the adult brain. Mol Psychiatry 7:140–156

8. Lemkine GF, Raj A, Alfama G, Turque N, Hassani Z, Alegria-Prevot O,Samarut J, Levi G, Demeneix BA 2005 Adult neural stem cell cycling in vivorequires thyroid hormone and its � receptor. FASEB J 19:863–865

9. Montero-Pedrazuela A, Venero C, Lavado-Autric R, Fernandez-Lamo I, Garcia-Verdugo JM, Bernal J, Guadano-Ferraz A 2006 Modulation of adult hip-pocampal neurogenesis by thyroid hormones: implications in depressive-likebehavior. Mol Psychiatry 11:361–371

10. Venero C, Guadano-Ferraz A, Herrero AI, Nordstrom K, Manzano J, deEscobar GM, Bernal J, Vennstrom B 2005 Anxiety, memory impairment, andlocomotor dysfunction caused by a mutant thyroid hormone receptor �1 canbe ameliorated by T3 treatment. Genes Dev 19:2152–2163

11. Iniguez MA, Rodriguez-Pena A, Ibarrola N, Morreale de Escobar G, BernalJ 1992 Adult rat brain is sensitive to thyroid hormone. Regulation of RC3/neurogranin mRNA. J Clin Invest 90:554–558

12. Oppenheimer JH, Coulombe P, Schwartz HL, Gutfeld NW 1978 Nonlinear(amplified) relationship between nuclear occupancy by triiodothyronine andthe appearance rate of hepatic �-glycerophosphate dehydrogenase and malicenzyme in the rat. J Clin Invest 61:987–997

13. Escobar-Morreale HF, del Rey FE, Obregon MJ, de Escobar GM 1996Only the combined treatment with thyroxine and triiodothyronine ensureseuthyroidism in all tissues of the thyroidectomized rat. Endocrinology 137:2490 –2502

14. Goumaz MO, Schwartz H, Oppenheimer JH, Mariash CN 1994 Kinetic modelof the response of precursor and mature rat hepatic mRNA-S14 to thyroidhormone. American J Physiol 266:E1001–E1011

15. Zavacki AM, Ying H, Christoffolete MA, Aerts G, So E, Harney JW, ChengSY, Larsen PR, Bianco AC 2005 Type 1 iodothyronine deiodinase is a sensitivemarker of peripheral thyroid status in the mouse. Endocrinology 146:1568–1575

16. Livak KJ, Schmittgen TD 2001 Analysis of relative gene expression data usingreal-time quantitative PCR and the 2[-��C(T)] method. Methods (San Diego)25:402–408

17. Team RDC 2005 R: a language and environment for statistical computing.Vienna, Austria: R Foundation for Statistical Computing

18. Carey VJ, Gentry J, Whalen E, Gentleman R 2005 Network structures andalgorithms in Bioconductor. Bioinformatics 21:135–136

19. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, EllisB, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, IrizarryR, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G,Tierney L, Yang JY, Zhang J 2004 Bioconductor: open software developmentfor computational biology and bioinformatics. Genome Biol 5:R80

20. Diez D, Alvarez R, Dopazo A 2007 Codelink: an R package for analysis of GEhealthcare gene expression bioarrays. Bioinformatics 23:1168–1169

21. Irizarry R, Gautier L, Bolstad B, Miller C, Astrand M, Cope LM, GentlemanR, Gentry J, Halling C, Huber W, MacDonald J, Rubinstein B, Workman C,Zhang J 2006 Affy: methods for Affymetrix oligonucleotide arrays. Availableat http://bioconductor.org/packages/2.2/bioc/html/affy.html

22. Gentleman R, Carey V, Huber W, FH 2007 Genefilter: methods for filteringgenes from microarray experiments. Available at http://bioconductor.org/packages/2.2/bioc/html/genefilter.html

23. Smyth GK, Michaud J, Scott HS 2005 Use of within-array replicate spots forassessing differential expression in microarray experiments. Bioinformatics21:2067–2075

24. Hochberg Y, Benjamini Y 1990 More powerful procedures for multiple sig-nificance testing. Stat Med 9:811–818

25. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U,Speed TP 2003 Exploration, normalization, and summaries of high densityoligonucleotide array probe level data. Biostatistics 4:249–264

26. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP,Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L,Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM,Sherlock G 2000 Gene ontology: tool for the unification of biology. The GeneOntology Consortium. Nat Genet 25:25–29

27. Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, Daverman R, DiemerK, Muruganujan A, Narechania A 2003 PANTHER: a library of protein familiesand subfamilies indexed by function. Genome Res 13:2129–2141

28. Cleveland WS, Grosse E, Shyu WM 1992 Statistical models in S. Chambers J,Hastie T, eds. Dordrecht: The Netherlands: Chapman and Hall, Kluwer Ac-ademic Publishers; 309–376

29. Sohl G, Maxeiner S, Willecke K 2005 Expression and functions of neuronalgap junctions. Nat Rev Neurosci 6:191–200

30. Brotchie JM 2003 CB1 cannabinoid receptor signalling in Parkinson’s disease.Curr Opin Pharmacol 3:54–61

31. Asua T, Bilbao A, Gorriti MA, Lopez JA, Del Mar Alvarez M, Navarro M,de Fonseca FR, Perez-Castillo A, Santos A 2008 Implication of the endocan-nabinoid system in the locomotor hyperactivity associated with congenitalhypothyroidism. Endocrinology 149:2657–2666

32. von Bohlen und Halbach O 2005 The renin-angiotensin system in the mam-malian central nervous system. Curr Protein Pept Sci 6:355–371

33. Suwelack D, Hurtado-Lorenzo A, Millan E, Gonzalez-Nicolini V, WawrowskyK, Lowenstein PR, Castro MG 2004 Neuronal expression of the transcriptionfactor Gli1 using the T�1�-tubulin promoter is neuroprotective in an exper-imental model of Parkinson’s disease. Gene Ther 11:1742–1752

34. Kato M, Seki N, Sugano S, Hashimoto K, Masuho Y, Muramatsu M, KaibuchiK, Nakafuku M 2001 Identification of sonic hedgehog-responsive genes usingcDNA microarray. Biochem Biophys Res Commun 289:472–478

35. Eisenhofer G 2001 The role of neuronal and extraneuronal plasma membranetransporters in the inactivation of peripheral catecholamines. Pharmacol Ther91:35–62

36. Wong HK, Sakurai T, Oyama F, Kaneko K, Wada K, Miyazaki H, KurosawaM, De Strooper B, Saftig P, Nukina N 2005 � Subunits of voltage-gatedsodium channels are novel substrates of �-site amyloid precursor protein-cleaving enzyme (BACE1) and �-secretase. J Biol Chem 280:23009–23017

37. Desplats PA, Kass KE, Gilmartin T, Stanwood GD, Woodward EL, Head SR,Sutcliffe JG, Thomas EA 2006 Selective deficits in the expression of striatal-enriched mRNAs in Huntington’s disease. J Neurochem 96:743–757

38. Quraishe S, Asuni A, Boelens WC, O’Connor V, Wyttenbach A 2008 Ex-pression of the small heat shock protein family in the mouse CNS: differentialanatomical and biochemical compartmentalization. Neuroscience 153:483–491

39. Ogden AT, Nunes I, Ko K, Wu S, Hines CS, Wang AF, Hegde RS, Lang RA1998 GRIFIN, a novel lens-specific protein related to the galactin family. J BiolChem 273:28889–28896

40. Kuespert K, Pils S, Hauck CR 2006 CEACAMs: their role in physiology andpathophysiology. Curr Opin Cell Biol 18:565–571

41. Helfand BT, Mendez MG, Pugh J, Delsert C, Goldman RD 2003 A role forintermediate filaments in determining and maintaining the shape of nervecells. Molecular Biol Cell 14:5069–5081

42. Konishi H, Namikawa K, Shikata K, Kobatake Y, Tachibana T, Kiyama H2007 Identification of peripherin as a Akt substrate in neurons. J Biol Chem282:23491–23499

43. Guo L, Lobenhofer EK, Wang C, Shippy R, Harris SC, Zhang L, Mei N, ChenT, Herman D, Goodsaid FM, Hurban P, Phillips KL, Xu J, Deng X, Sun YA,Tong W, Dragan YP, Shi L 2006 Rat toxicogenomic study reveals analyticalconsistency across microarray platforms. Nat Biotechnol 24:1162–1169

44. Shi L, Tong W, Fang H, Scherf U, Han J, Puri RK, Frueh FW, Goodsaid FM,Guo L, Su Z, Han T, Fuscoe JC, Xu ZA, Patterson TA, Hong H, Xie Q, Perkins




RG, Chen JJ, Casciano DA 2005 Cross-platform comparability of microarraytechnology: intra-platform consistency and appropriate data analysis proce-dures are essential. BMC Bioinformatics 6(Suppl 2):S12

45. Shi L, Reid LH, Jones WD, Shippy R, Warrington JA, Baker SC, CollinsPJ, de Longueville F, Kawasaki ES, Lee KY, Luo Y, Sun YA, Willey JC,Setterquist RA, Fischer GM, Tong W, Dragan YP, Dix DJ, Frueh FW,Goodsaid FM, Herman D, Jensen RV, Johnson CD, Lobenhofer EK, PuriRK, et al. 2006 The MicroArray Quality Control (MAQC) project showsinter- and intraplatform reproducibility of gene expression measurements.Nat Biotechnol 24:1151–1161

46. Koritschoner NP, Alvarez-Dolado M, Kurz SM, Heikenwalder MF, HackerC, Vogel F, Munoz A, Zenke M 2001 Thyroid hormone regulates the obesitygene tub. EMBO Rep 2:499–504

47. Thomas EA, Danielson PE, Sutcliffe JG 1998 RGS9: a regulator of G-proteinsignalling with specific expression in rat and mouse striatum. J Neurosci Res52:118–124

48. Denver RJ, Ouellet L, Furling D, Kobayashi A, Fujii-Kuriyama Y, PuymiratJ 1999 Basic transcription element-binding protein (BTEB) is a thyroid hor-mone-regulated gene in the developing central nervous system. Evidence fora role in neurite outgrowth. J Biol Chem 274:23128–23134

49. Haas MJ, Mreyoud A, Fishman M, Mooradian AD 2004 Microarray analysisof thyroid hormone-induced changes in mRNA expression in the adult ratbrain. Neurosci Lett 365:14–18

50. Morte B, Iniguez MA, Lorenzo PI, Bernal J 1997 Thyroid hormone-regulatedexpression of RC3/neurogranin in the immortalized hypothalamic cell lineGT1–7. J Neurochem 69:902–909

51. Pipaon C, Santos A, Perez-Castillo A 1992 Thyroid hormone up-regulatesNGFI-A gene expression in rat brain during development. J Biol Chem 267:21–23

52. Alvarez-Dolado M, Gonzalez-Moreno M, Valencia A, Zenke M, Bernal J,Munoz A 1999 Identification of a mammalian homologue of the fungal Tom70mitochondrial precursor protein import receptor as a thyroid hormone-reg-ulated gene in specific brain regions. J Neurochem 73:2240–2249

53. Thompson CC 1996 Thyroid hormone-responsive genes in developing cere-bellum include a novel synaptotagmin and a hairless homolog. J Neurosci16:7832–7840

54. Vargiu P, Morte B, Manzano J, Perez J, de Abajo R, Gregor Sutcliffe J, BernalJ 2001 Thyroid hormone regulation of rhes, a novel Ras homolog gene ex-pressed in the striatum. Brain Res Mol Brain Res 94:1–8

55. Dong H, Yauk CL, Williams A, Lee A, Douglas GR, Wade MG 2007 Hepaticgene expression changes in hypothyroid juvenile mice: characterization of anovel negative thyroid-responsive element. Endocrinology 148:3932–3940

56. Quignodon L, Grijota-Martinez C, Compe E, Guyot R, Allioli N, LaperriereD, Walker R, Meltzer P, Mader S, Samarut J, Flamant F 2007 A combinedapproach identifies a limited number of new thyroid hormone target genes inpost-natal mouse cerebellum. J Mol Endocrinol 39:17–28

57. Coulombe P, Schwartz HL, Oppenheimer JH 1978 Relationship between theaccumulation of pituitary growth hormone and nuclear occupancy by triiodo-thyronine in the rat. J Clin Invest 62:1020–1028

58. Maxeiner S, Kruger O, Schilling K, Traub O, Urschel S, Willecke K 2003Spatiotemporal transcription of connexin45 during brain development resultsin neuronal expression in adult mice. Neuroscience 119:689–700

59. Cirelli C, Gutierrez CM, Tononi G 2004 Extensive and divergent effects ofsleep and wakefulness on brain gene expression. Neuron 41:35–43

60. Niculescu 3rd AB, Segal DS, Kuczenski R, Barrett T, Hauger RL, Kelsoe JR2000 Identifying a series of candidate genes for mania and psychosis: a con-vergent functional genomics approach. Physiol Genomics 4:83–91

61. Stansberg C, Vik-Mo AO, Holdhus R, Breilid H, Srebro B, Petersen K,Jorgensen HA, Jonassen I, Steen VM 2007 Gene expression profiles in rat braindisclose CNS signature genes and regional patterns of functional specialisa-tion. BMC Genomics 8:94

62. Greengard P 2001 The neurobiology of slow synaptic transmission. Science294:1024–1030

63. Vargiu P, De Abajo R, Garcia-Ranea JA, Valencia A, Santisteban P, CrespoP, Bernal J 2004 The small GTP-binding protein, Rhes, regulates signal trans-duction from G protein-coupled receptors. Oncogene 23:559–568

64. Gerendasy DD, Sutcliffe JG 1997 RC3/neurogranin, a postsynaptic calpacitinfor setting the response threshold to calcium influxes. Mol Neurobiol15:131–163

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.




thyroid hormone action in the adult brain: gene …digital.csic.es/bitstream/10261/22145/1/thyroid...

Documents